But I now move on to what Milton Friedman calls the marketing problem, because this is very important to me. Imperfections of the political market

Transcription

1 Discussion of Inflation and Government, by Samuel Brittan by Milton Friedman In Inflation: Causes, Consequences, Cures, pp Institute of Economic Affairs. First published by the Institute of Economic Affairs, London, I begin with a mea culpa. We must all learn from our mistakes. I supported the 1972 wage-price freeze, though not the subsequent phases 2 and 3, very much on the lines of the arguments from the Argentine case presented by Professor Milton Friedman. I thought it would be helpful in puncturing inflationary expectations in conjunction with, not a deflation, you must not use that word, but an anti-inflationary monetary and fiscal policy. Of course I should have realized more, after all these years, about the characters of the people enforcing it. It was in fact used, not as an accompaniment to sound monetary and fiscal policy, but as an excuse for pumping more money into the economy in the pursuit of so-called full employment. The result will probably be to land us eventually with the highest post-war rate of unemployment we have ever seen. But I now move on to what Milton Friedman calls the marketing problem, because this is very important to me. Imperfections of the political market If it were not so late on in his visit, I would be very interested to hear his ideas for improving the political market. People talk a great deal about the imperfections of the commercial market, but the most imperfect market we have is the political market. This is the real problem, even more than inflation, that threatens western democracy at the moment. I think that people who are called monetarists sometimes give an unnecessarily arid and misleading impression of their own case by the way in which it is often presented in popular discussion. They give the impression that it is something rather technical to do with the money supply and the Bank of England should be left to get on with it. It is made to seem like a very technical argument about M1 and M3 and disintermediation. First of all, the most important proposition relates to money expenditure. The view that the most important influence on money expenditure is the money supply is, I think, valid, but most of what we are saying would be equally true if it turned out empirically that budgetary policy was more important. Second, there is something even more important than that. (This is why I think the label monetarist is so awful and the label neo-classical is even more awful, but I wish I could think of a better one.) The central point that sound economists are making is that there is no easy way by which the government can create a lower rate of unemployment than is allowed by the workings of the economy with all its imperfections, by the state of the labor market, and by all the other real forces at work. We are too much influenced by the situation in the 1930s when what was required was an increase in the money supply instead of a reduction in the level of money wages. This is not the situation now. Mass unemployment scare 1

2 The central point of the position and I am putting this across really for the benefit of the skeptics in this room who have not been as vocal as they should be the central point of what is miscalled the monetarist position, is that we cannot spend our way into as low a rate of unemployment as we would like. The so-called monetarists are not advocating mass unemployment. Their argument in the current UK context is that we are going to get the unemployment in any event. The question is whether we get it now at a still moderate rate of inflation or whether we get it later at a state of hyper-inflation, when possibly, owing to the breakdown of the monetary system, we will have a real depression and not a recession. What we are saying is that you do not avoid the unemployment that you are going to get by the normal U- turns and by printing money. Most of us say, whenever we are confronted with people like Milton Friedman and David Laidler, that we agree with 90 per cent of what they say. I think I know them well enough to say that this is more irritating to them than to have out-and-out opponents. I remember my tutorials with Milton Friedman many years ago and I know how that remark irritates him. But most of us, being human, are in that position people do not agree with each other 100 per cent. Now 5 per cent of my queries (rather than disagreements) are of a kind which can only be discussed in what would really have to be a tutorial. But the other 5 per cent relate to what could be termed the political marketing of what one might call the market economy case. Unfortunately there is mainly a market for this case only in the Conservative party, perhaps a tiny fraction of the Liberal party and an even tinier fraction of the Labour party. Anybody who knows my political record will know that it does not give me any pleasure to put it that way, but that is the position. Funny alliance? The impression gained is that there is a funny kind of alliance between the extreme left and the extreme right in saying that unions are wonderful. I think too that, in the heat of the moment, a good many Conservatives and others, in an understandable reaction to the policies of the last Conservative Government in the past couple of years, which attempted to substitute price and incomes control for a sensible economic policy, have made the mistake of almost whitewashing the unions. Anybody who has listened to Milton Friedman and read carefully what he has written, will realize that the unions do a great deal of damage. They create unemployment; they deprive people, usually poor people, of job opportunities; they price the most handicapped and the most unfortunate people either out of the labor market or into worse employment than they could otherwise get. And as we are tending to move into a new medievalism, into a new world of so-called just but really very unjust prices, it might be worth pointing out, in medieval terms, that the unions are not the Robin Hoods they are supposed to be on every television program. They are the robber barons of the system. And this is true even if they have zero effect on the rate of inflation. I would like to argue and develop this a bit more in later discussion, but perhaps the best way of stating the case is to say that, if one takes a dynamic model which I might try and develop, the effects of unions in the real world are probably to raise the natural rate of unemployment rather than immediately to raise the rate of inflation. Of course governments do not like the natural rate of unemployment and from that follows all the consequences we have been talking about. 2

3 I believe in a social market economy, primarily for ethical reasons and because of a belief in personal freedom rather than because of anything to do with maximization of the GNP. It is unfortunate that, in the way the political debate has focused in this country, the whole case for a social market economy, everything represented by those economists who are still prepared to believe that relative prices matter, has become over-simplified into the word monetarism. I mix with people in the twilight world between politics and economics and they all say to me, I wish you people would not talk exclusively about the money supply. I sometimes quote to them a remark of Milton Friedman s which in turn derives from John Stuart Mill: The more I learn about money the more I realize how little it matters. But perhaps to get to a world in which we can forget about the money supply and talk about the real problems of resource allocation and genuine ways of alleviating poverty, as distinct from phony and deceptive methods such as rent controls and housing subsidies, we may have to get the inflationary psychosis out of the system. Indexation and a genuine floating rate Indexation has come up in basically two contexts. One is as an adjunct to a policy of gradually decelerating the growth of the money supply and as a way of alleviating the unemployment consequences of so doing. If I were more politically optimistic I might leave it there. But I would still be prepared to support indexation, not just as an adjunct to a cure, but also as a method for enabling some semblance of civilized life to continue during what could be generations of currency disorders. There is an analogy here in the case for floating rates, only unfortunately nothing in the world seems to be able to induce the Bank of England to operate a genuine floating rate. By that I do not mean clean floating, I do not mean no intervention; I mean simply allowing market forces to influence the rate. We have had all the guarantees Milton Friedman has been talking about, plus many more that are secret, plus a good deal of public sector borrowing with no forward risk on the part of the borrower because the risks have all been taken by the central government. This is not any kind of floating rate, not even a dirty floating rate. But taking the general international scene, I am quite sure that the adjustments to the real changes, such as the sudden eruption of the oil cartel, would have been far more difficult if we were still in a world of fixed rates. I now come to the main analytical question which I would like to ask our distinguished Chairman and our monetarists. My friend Peter Jay is in the habit he is not exactly a colorless character of saying that medium-term stabilization policy would involve tolerating unemployment in the low millions. Now there are basically three unemployment rates we have to consider. First, there is the wishful thinking unemployment rate, the unemployment rate which exists either in White Papers or in supposedly secret drawers in Treasury offices, the unemployment rate that politicians or officials think they could achieve. We can, I think, forget about that. Second, there is the so-called natural rate of unemployment. Perhaps we ought to call it the equilibrium or sustainable rate of unemployment, which without institutional reform a system can settle down to at a constant rate of price change. 3

4 Third, there is the long-term question we ought to be discussing: how we can bring that rate down. Fourth, there is the rather horrible question of the transitional rate of unemployment. Assume you want to go down from an underlying rate of inflation of 20 per cent to one of 10 per cent over four years, leaving out ripples on the route. The question one is always being asked by socalled practical people is, What will be the transitional unemployment? It is quite understandably very difficult to get an answer. But the question is this, if we were to attempt to mitigate the hardship involved, say, by focusing government expenditure on the depressed areas, on the areas particularly hard-hit, would this be helpful? I read a column in Newsweek by Professor Friedman in which he was very skeptical of the idea of special help for special areas. But suppose one puts this in the context of a budgetary ceiling and a determination not to finance any budgetary deficit by creating money, which in practice means a very low budgetary deficit. The question is not what the total size of the budget should be, but whether it should be spent in areas, say, in the Southeast where employment is over-full, or whether it should be spent in areas which have regional problems where industries have declined, and where, given the rigidities of the wage structure, people cannot find work. If you switch your expenditures, switch given totals of expenditure towards hard-hit areas or towards certain industries, might you not be able to get a lower rate of transitional unemployment for the same effect on inflation? These are quite genuine, not rhetorical, questions. DISCUSSION LAIDLER: Mr. Brittan raised a question of how much unemployment to reduce the inflation rate how fast. We do not know very much about it but we do have one observation which is the period when the unemployment rate peaked out at that magic number one million. I have the consumer price index in front of me in percentage rate of change terms. From the first/second quarter 1971 inflation rate in this country (retail prices) was 11-9 per cent. From the first/second quarter 1972 it was 4-6 per cent. So you can more than halve that kind of inflation rate in a year with a million unemployed. And that is a very rapid decrease in the rate of inflation. I do not know, however, how far we can extrapolate from that experience to the next. FRIEDMAN: There are many other cases. This is an important subject for study, and I hope there are some potential PhD students here because I think we need half a dozen dissertations on various episodes which could give us a better handle on this number. Let me give you some of the examples. George Schwartz referred to the greenback period, , when the price level was cut in half, unemployment was trivial, the fastest rate of growth of any decade in American history. That is a favorable case, although at the time there was a lot of yelling. France, 1925: a very effective, rapid stabilization in which the inflation rate was reduced sharply, very little unemployment. Britain in the 1920s, a bad case, needed a 10 per cent decline in prices. Result, very high unemployment for a long period. You have dozens of cases in South America and lots of other episodes. In many of those I know, you have had a very substantial reduction in inflation at very low cost. 4

5 LORD ROBBINS: What was the strength of the trade union movement in the greenback period, in France in the period you mentioned, and in this country in the 1920s? FRIEDMAN: In the US it was negligible. I don t know about France. LORD ROBBINS: It was pretty weak. FRIEDMAN: It may well have been. In Britain in the 1920s it was pretty strong. I am not quarrelling with the view that the existence of a trade union may make this problem more difficult than it otherwise would be. One of the things we would hope would come from a serious study of the kind I am describing is how important these various factors are. But I want to emphasize it is not only trade unions that matter. Let me give you the comparison between the US, 1873 to 1879 and 1929 to You had essentially the same decline in nominal income in the two periods. The first was accompanied by rapid economic growth, a sharp fall in prices, and very little fall in employment. The second was accompanied by a sharp decline in employment, and a much slower decline in prices, much more unemployment. Yet trade unions were not very strong in the United States in So I think the phenomena at work involve more than trade unions. JOHN FLEMMING: I can offer a policy possibly superior to some of the versions of monetarism but liable to attract a label which would certainly be worse. If the risks, in terms of unemployment, associated with the kind of monetary policy which some people are advocating are very great, in the sense that they would commit themselves to a tight monetary policy although not knowing whether the resultant unemployment in 12 months time was going to be half a million or one and a half million, it might be very substantially preferable to have a policy which was designed to bring about a certain very modest rate of increase in unemployment itself. Such a policy might have considerable advantages, given the ignorance that we are in about the effects of certain policies, over the adoption of a monetary rule. But it seems to me quite clear that, however modest one were to make the planned increase, one would receive even more abuse for advocating such a policy than the abuse that is already poured on the heads of monetarists. LORD ROBBINS: Now I think the time has come to ask such of the speakers as are present, except the chairman, to make further observations on what has been raised in the course of the discussion. LAIDLER: I would like to make two or three points. First, about wage and price controls: the standard case for them in this country has been that they affect expectations. That really is a very spurious argument, because any economic policy that is believed in will affect expectations. There is nothing unique about wage and price controls. And I would have thought, honestly, in the present situation that if there was one economic policy that nobody is going to believe in, when it is enforced, it is another bout of wage and price controls. Second, there is some very simple economics that tells you why wage and price controls do not work. The market price is a price agreed between a buyer and a seller. Effective price control fixes a price below market price. That means there are buyers who are willing to pay a higher price and sellers who are willing to accept a higher price. Therefore you have the open 5

6 possibility and incentive to collusion on both sides of the bargain. What then happens is that the least law-abiding among us start engaging in that kind of collusion. And the more law-abiding among us unfortunately are put in the position of either having to go without or break the law. This was brought home to me very strongly during the Health freeze by a neighbor who has a number of factories. He told me, Quite frankly, next week either we start breaking the law like our competitors or we start closing down our factories. And that is appalling, absolutely appalling. I think we should remember what wage and price controls do to business morality and respect for the law. Third, the more technical business about what you might want to do in order to bring the rate of monetary expansion under control. We have a set of monetary institutions in this country that go under the title Competition and Credit Control. Lately there has been quite a stirring in the press to the effect that Competition and Credit Control is a monetarist invention and the one place where the monetarists went wrong. Patrick Cosgrave said exactly that in The Spectator a few weeks ago. The problem of Competition and Credit Control is that it does not have enough of either. It does not have enough competition because the building societies were explicitly given a privileged position. A great deal of the monetary expansion in the last few years in this country has come through attempts to use monetary policy to keep interest rates down and protect owner-occupiers. Now of course it has not protected owner-occupiers; it has put them much further out on a limb than they otherwise would have been; and they are all in a much worse position now than they would have been if this policy had never been followed in the first place. There are people now whose entire equity in their houses has been wiped out in the last 12 months, who are faced with cash flow problems with their mortgages, and who just cannot cope. As for the credit control side of it, we have a system Heaven help us! of reserve assets that includes short-term government securities. This means that having got a good short-term government securities market, which gave us the chance of insulating monetary policy from the government borrowing requirement, we have thrown away that advantage. The second thing we have is a system in which the banking system can actually manufacture its own reserve assets by lending at call to the discount market. I do not believe that it is possible to conduct the kind of monetary policy that I would advocate, and Milton Friedman would advocate, while we have that institutional set-up. I do not understand why we cannot have a simple rule that says, Commercial banks must hold cash or deposits with the Bank of England at least equal to 10 per cent or 18 per cent or 15 per cent of their deposit liabilities, and leave it at that. MILTON FRIEDMAN: There are a number of comments I would like to make. One goes back to Lionel Robbin s earlier discussion of three propositions. The first one was: do we agree that the control of the credit base is an indispensable condition of coping with inflation? We all said yes. The second one was: inflation is always caused by failure to observe prescription number one. I want to make a comment on that. I have for years been trying to collect examples or counter-examples to that proposition. I have almost never addressed a group of economists without asking them if any of them had one, and I have myself discovered one and only one counter-example. That is the Korean War inflation in the United States in , which is the one case I know of a pretty substantial rate of price rise which cannot be attributed to a prior excessive expansion of the quantity of money, but rather was due to an autonomous explosion in 6

7 the velocity of circulation. Now it is an interesting exception because, to put that old aphorism correctly, it s an exception that tests the rule. It was very brief because it was not supported by monetary expansion. And the rate of inflation in the United States came down from something like 15 or 20 per cent in wholesale prices to something like zero per cent within six months, because there was no support for this autonomous explosion in velocity. I do not know of any other example. I think it would be interesting to collect all such other examples that anybody can find. I have brought this up in order to express my usual plea. If there is anybody who knows of any case that could with any plausible reason be regarded as an exception, I would appreciate very much if he would let me know. LORD ROBBINS: This is a convenient moment to make this point. There are almost endless ambiguities in social and historical explanations aroused by the term cause. I am not quite sure whether you were here when Professor Coats indulged in his very entertaining historical reflections. It seemed to me then that historians whose views he was drawing our attention to were people who were prepared to or anxious to take things further back as the sort of explanation which we as economists would think appropriate. And one can argue almost indefinitely about semantics. Bob Coats mentioned some historic cases of inflation, the silver inflation of the Renaissance and the German inflation, and so on, and the varying explanations given by historians. Clearly the historians were looking for something other than what economists would react to. He did not mention the episode of the Assignats which I think is extraordinarily interesting from our point of view. You had a revolution, you had people with new ideas of what government should do. They went all out for a kind of money which was bound to break down by indefinite multiplication. Now we, as economists, would say that a sufficient explanation is that they did issue the money. But supposing an historian should say, Oh, about the cause of that, I am improvising entirely, I am not expert at all in the details of the history. The cause was that some of the Jacobins can be shown to have had very unfortunate experiences with their parents, and consequently they took a jaundiced view of certain issues. Thus when somebody suggested to them that certain things should be done by very unorthodox policy they said, Oh, what the hell, let it go forward. Now I suppose we would all agree that, if that were true (it probably isn t), it would be a cause. And yet I do not think economists would want to weave that into their theory of the causes and conditions of inflation. Now I come back to our little controversy. We all, I think, agreed with you there hasn t been a dissident voice that in 9 cases out of 10 when a condition of inflation persists, there is an undue expansion of expenditure, expenditure gingered up by an increased credit base. Where we disagreed, where there was a good deal of talk earlier, was to what extent you could think of possible kinds of behavior within the economic system which could, so to speak, be combined in the conception of causation with the failure of the government to do its duty to maintain the standard of value. And that, in spite of all that has been said, I do believe in the end is a bit of a semantic question. Although you and David Laidler adduced evidence to persuade us that a great deal of what is said against trade union action is based on misapprehension, that some trade union action at any rate is vainly trying to keep up with relatives and the price level, you admitted to me that you knew you could have a slow explosion. And in conversation with Peter Jay we agreed that, if people got it into their heads that they were always going to ask for 105 per cent of the GNP, that would cause an indefinite expansion towards infinity which could only be stopped by the substitution of new money. 7

8 So are we really so much in disagreement? If I said that if, in the process of putting on the brakes on the rate of increase of money supply, some miracle occurred and the leaders of the Trades Union Congress passed a self-denying ordinance not during that period to do anything to maintain their position as they considered it ought to be, I do not think you would disagree that probably there would be rather less unemployment than there would be if they were in a frightfully aggressive mood. So where do we part company? FRIEDMAN: I myself try to avoid the use of the word cause for the very reasons you cite; it is a tricky and unsatisfactory word. When I use it in this connection I always speak of the change in the money supply as a proximate cause and say that the deeper causes must be found in what are the explanations for the rise in the money supply. And yet I am not persuaded that the difference between us in these arguments is purely semantic. I believe there is a very real difference. It is brought out by your final case. Somehow or other the problem I get into is to try to distinguish analytical issues from practical issues. The final statement sounds to me like the following: if we may provisionally assume that water runs uphill, then will you agree with me that we will get more water to the top of the hill than otherwise. LORD ROBBINS: No, I am not asking that. I am asking whether, if the conditions that prevailed in your Argentinian case were to occur in other times and in other places, there would not be an easier passage than if the reverse conditions prevailed. FRIEDMAN: I agree they would. But then we have to face up to the analytical and empirical issues. Under what circumstances can they prevail and under what circumstances do they prevail? I believe that it is an illusion of the worst kind to suppose that they always prevail, as both Brittan and Laidler have said. Suppose that this so-called income and price freeze is an example of the kind of thing you are describing. I think it was possible in Argentina to work it, partly because the rate of inflation was so extremely high, and partly because a government was able to persuade people that they were really going to conduct this policy, and as an adjunct to that it may have had some psychological effect. Incidentally, trade unions were weak in your sense, not in the political sense but in an economic sense. If you could recreate those situations here it might have some effect. But I think in the real, concrete world it would not. Therefore there is a real difference. But there are a couple of other points I would like to make, one deriving from David Laidler and one from Sam Brittan. You have the interesting phenomenon that whereas David Laidler came to Chicago, Chicago came to Sam Brittan. I must say there are few things that annoy me more, just as they annoy David Laidler, than the continual repetition of black box arguments. What is the mechanism? I have decided that the right way to answer those questions is simply to invite people to read David Hume s essay on money, because we must not think that we are discovering very many new things. About a year or so ago I was asked by the Bell Laboratories in New Jersey to talk to their scientists not economists, but scientists of various kinds about what we knew about money as part of a program of bringing scientific contributions to their notice. In the course of preparing the talk I re-read David Hume s essay. And I must say I was in one respect very much depressed by it. There was so little I knew that was not there. We have not learned very much in two centuries. If anybody asks what is the mechanism whereby an increase in the money supply brings about an increase in prices, what David Hume has to say answers that question about as well as anything else I know. I ask myself, What do we really 8

9 know that he did not know? and there is only a very little. We know the numbers better; we can attach numerical magnitudes; and we are a bit more sophisticated about the dynamic process of decelerating and accelerating inflation than he was; but beyond that he had it all. Second, I recently re-read William Stanley Jevon s essay on the serious fall in the value of gold. And lo and behold I was shocked and again depressed by finding something I thought I had discovered. Jevons said an increase in currency is followed within a space of two years by an increase in prices. That was the lag he found in 1860.That is the lag I found in the United States. It s a lag that s true in Britain today. So that despite the increase in the speed of communication and in financial sophistication and so on, apparently the British and American lag and the adjustment of prices to an increase in the supply of money has remained unchanged. LORD ROBBINS: You re not prepared to bet large sums of money on the continuance of that exact lag? FRIEDMAN: No, that is where we know a little more. We know that the length of that lag depends on the rates of price inflation. In a world where people have been accustomed to relatively slow changes in prices, the two-year lag will prevail. As people become accustomed to more rapid changes in the rate of inflation, the lag will shorten. The third historical reference I want to give you is J.E. Cairnes on the influence of the Australian gold discoveries. He did something there which has been notably absent. He asked: if you have an increase in gold, which prices will react first, which second, which later? What will be the influence on relative prices of a general inflation? The fascinating thing is that his argument then would not hold water today, but the results he arrived at then have been pretty good from then to now. So far as I know nobody in recent times has explored that issue, either empirically in a systematic way or theoretically. He divided commodities into vegetable, animal and manufactured products. His analysis stemmed from the fact that the increase in the quantity of money in the first instance came into the hands of consumers, whereas today all of our analysis proceeds on the assumption that it first goes into the credit market so that the initial first round effect is wholly different, and yet the results seem to have been the same. Finally, on Sam Brittan s question about mitigating hardship. I agree completely that the same total budget can be spread in ways which will alleviate hardship more or less. The policies I was writing about (in Newsweek) in the United States were not regional, but so-called public unemployment policies. And I submit to him that the policies that will be most effective to mitigate hardship will be improvements in unemployment measures for long-term unemployed. Avoid like the plague trying to do very much to improve the conditions for short-term unemployment because that is where you have the biggest leakage and the biggest extent to which you will increase the natural rate of unemployment. But if you take the very long-term Hardship case, then I think there is a good deal to be said for mitigating it. I am very skeptical about the regional unemployment measures for doing it because it seems to me that much of that will be wasted in shifting people who are employed rather than in assisting the people who most need assistance. 4/17/13 9